Ca-and Mg-rich waste as high active carrier for chemical looping gasification of biomass

Xin Niu,Laihong Shen

1 School of Energy and Power Engineering,Nanjing University of Science and Technology,Nanjing 210094,China

2 Key Laboratory of Energy Thermal Conversion and Control of Ministry of Education,School of Energy and Environment,Southeast University,Nanjing 210096,China

Keywords:Biomass Syngas Gasification Ca-and Mg-rich waste Chemical looping

ABSTRACT Chemical looping gasification (CLG) is a promising technology for high-quality syngas production.One key issue to successful CLG is the selection of high-performance oxygen carrier.In this study,several Ca-and Mg-rich steelmaking wastes from steel industry,such as blast furnace slag(BF slag),blast furnace dust(BF dust)and Linz-Donawitz converter slag(LD slag),were used as oxygen carriers in chemical looping gasification of biomass.The results showed that the reducibility of Ca-and Mg-rich waste,especially LD slag and BF dust,was superior to that of hematite.Considering long-term operation,the cyclic stability of steelmaking waste was tested.BF dust showed a poor stability,while the other carrier (hematite,BF slag or LD slag) presented an excellent stability during multiple redox cycles in spite of partial sintering and agglomeration.Moreover,the effects of supply oxygen coefficient (O/B ratio) and reaction temperature on CLG of biomass were investigated.The results revealed that Ca-and Mg-rich waste exhibited a higher syngas production compared to hematite.The higher performance could be attributed to the improved reduction rate of Fe2O3 and gasification rate of biomass by Ca or Mg in steelmaking waste.In addition,LD slag exhibited the higher gas value at the O/B ratio of 1 at 900 °C.As a consequence,LD slag was an appropriate oxygen carrier for CLG of biomass in terms of perfect reducibility,superior cyclic stability and high reactivity.

1.Introduction

Biomass,a renewable resource with CO2neutrality,shows the potential to generate value-added chemicals through thermochemical methods.Among the thermochemical methods,gasification is one of the most promising technology for biomass conversion [1].However,the inevitable tar production and coke formation will eventually lead to the poor gas quality and low gasification efficiency [2].In response to these difficulties,Chemical looping gasification(CLG)is proposed aiming for high quality syngas and possible tar cracking[3].In this process,the lattice oxygen in oxygen carrier is designed to fulfil the oxygen demand of gasification,in place of molecular oxygen.CLG consists of two interconnected fluidized bed reactors,an air reactor and a fuel reactor,with an oxygen carrier circulating between them.In the fuel reactor,the oxygen carrier provides lattice oxygen for fuel gasification and converts into a less oxidized form.Then the reduced oxygen carrier is transported to the air reactor for regeneration.Therefore,the design and selection of an appropriate oxygen carrier is essential to successfully operate a CLG system.The properties of oxygen carrier,such as sufficient reactivity,selectivity to H2and CO,high mechanical strength,low agglomeration tendency,and environmental friendly are the main factors to consider for a required oxygen carrier [3].

To date,the oxides of Fe,Ni,Cu,Mn and Co have been selected as the potential oxygen carrier materials.Among these carriers,Febased oxygen carrier is widely used for its cost-effectiveness,environmental sustainability and low degree de-fluidization problems[4–8].It is important to mention that iron oxide has been widely used as a catalyst in tar cracking and reforming [9–11].Additionally,iron oxide can promote water splitting reaction for hydrogen production [12].Nevertheless,Fe-containing oxygen carrier demonstrated moderate reaction reactivity in gasification with different fuels (such as coal [7,13],biomass [5,6],biomass char [14],microalgae[15],cellulosic[8],petroleum coke[16]).The moderate reactivity of Fe-based oxides might hinder the application development [17,18].To address this problem,metal ions are mixed with iron oxide via chemical or mechanical preparation method.On the one hand,it is possible to combine iron with a more active metal ion (such as Ni2+,Cu2+).Wei et al.[19] indicated that Fe-Ni oxide displayed a higher gasification efficiency of biomass compared to Fe2O3/Al2O3oxygen carrier.Niu et al.[20–21] investigated the reactivity of bimetallic Cu-Fe oxygen carrier with biomass.It was found that the synergistic reactivity of Cu-Fe oxide was achieved for tar decomposition during the biomass-derived CLG process.On the other hand,incorporation iron with alkali/alkaline metal ion (such as K+,Na+,Ca2+,Ba2+) could improve the reduction reactivity of Fe and enhance the lattice oxygen diffusivity[22–24].The research of Siriwardance et al.[25]pointed out that the rate of gasification and selectivity for synthesis gas production was significantly higher when calcium-ferrite (CaFe2O4) and barium-ferrite(BaFe2O4) were present during steam gasification of coal.Chan et al.[26] reported that Ca2Fe2O5achieved higher conversions of steam of 75% compared to 62% for 60 wt% Fe2O3-ZrO2.Similarly Sun et al.[27]pointed out a higher hydrogen yiled by using Ca2Fe2-O5as oxygen carrier.This could be due to the significant effect of Ca on the Fe3+reduction and Fe0oxidation.Work by Sun et al.[28] showed that the addition of Ca2Fe2O5resulted in a 13.4%increase in gas yield during pine wood gasification.Hu et al.[29]used Fe2O3/CaO oxygen carrier for CLG of biomass.Resluts showed that hydrogen production was promoted by one step reduction and oxidation of Ca2Fe2O5.

Considering that solid fuel has a high ash content which can lead to the agglomeration and deactivation of oxygen carriers,low-cost carriers are highly interesting.Steelmaking waste,generated in the manufacturing processes of pig iron and steel industry,consists of iron (Fe),calcium (Ca),magnesium (Mg),silicon (Si),aluminum (Al) and other elements [30].Because steelmaking waste has a relatively high Fe content,it has been of interest as an attractive candidate for chemical looping process.Xu et al.[31]investigated the performance of four low cost iron-based oxygen carriers (byproducts from the steel industry) in a batch fluidized bed reactor.It was reported that Bio_A and bio_C exhibited higher reaction rate and integral CO2yield with syngas compared to that of ilmenite,while LDst and AQS_US showed comparable performance to ilmenite.Moldenhauner et al.[32]reported that the flue gas consisted mostly of H2in CLG of black pellets using Linz-Donawitz slag as oxygen carrier,indicating this slag would be suitable for large-scale operation.Nevertheless,the lifetime of this material was estimated to be about 110–170 h.Additionally,Rydén et al.[33] showed that LD-slag was a very cheap and readily available oxygen-carrying bed material in oxygen carrier aided combustion (OCAC) of wood chips.St?rner et al.[34]investigated the interactions of iron mill scale and steel converter slag (LD-slag) with potassium ash model compounds.The results show that these materials should not be rejected as oxygen carriers for CLC based on potassium ash interaction.Hildor et al.[35]pointed out that Linz-Donawitz slag (LD slag) could be used as an active bed material in chemical looping process.However,the reactivity of LD slag towards syngas,CH4and C6H6was reduced as a function of time in the 12 MWthcirculating fluidized bed boiler.

Remarkably,few studies have reported the application of different kinds of steelmaking waste as oxygen carrier for CLG of biomass.In this study,blast furnace slag (BF slag),blast furnace dust (BF dust) and Linz-Donawitz converter slag (LD slag) were used as oxygen carriers for chemical looping.Hematite was also tested as a reference material.The reduction reactivity of these materials with gaseous fuels (H2,CO,CH4) was evaluated through TGA experiments.Additionally,the evolution of microstructure,cyclic stability and reaction mechanisms were evaluated.Furthermore,the functions of O/B and reaction temperature on gas yield in CLG of biomass were investigated in a batch fluidized bed reactor.

2.Experimental

2.1.Materials preparation

The tested oxygen carriers were blast furnace slag (BF slag),blast furnace dust(BF dust),Linz-Donawitz converter slag(LD slag)and hematite.All of these materials were supplied by Nanjing steelmaking factory.These materials were calcined at 950 °C in a muffle oven for 4 h to completely oxidize and increase the mechanical strength.The calcined particles were subsequently crushed and sieved to a particle size of 180–300 μm.The elemental compositions of the samples were analyzed using XRF and the results were shown in Table 1.

Rice husk,derived from Jiangsu province (China),was used as the biomass fuel in this work.The rice husk particles were crushed and sieved into the desired size-fraction,generally 600–850 μm.The proximate and ultimate analyses results were shown in Table 2.

Table 1Chemical compositions of the four oxygen carriers (%,mass)

Table 2Proximate and ultimate analyses of rice husk (%,mass)

2.2.Thermo-gravimetric analysis

The reduction reactivity experiments were conducted using a thermo-gravimetric analyzer (STA 449-F3 TGA,Netzsch).The effects of reducing agents(H2,CO and CH4)and reduction temperature (850,900 and 950 °C) on the oxygen carrier reactivity were investigated.During the experiments,the gas flow rate was maintained at 200 ml.min-1.In a typical experiment,around 100 mg oxygen carrier samples were loaded in a crucible and heated to the desired temperature in an air atmosphere.After stabilization,the system was purged with N2.Subsequently,reducing gas (5%H2in N2,15% CO in N2or 15% CH4in N2) was introduced to the reactor and maintained for 5 min.The use of a lower concentration of H2was to avoid the reduction of Fe2O3to Fe.

The oxygen transport capacity (ROC) is the mass fraction of the oxygen carrier being theoretically transferable as oxygen.

where moxis the mass of fully oxidized sample,mredis the mass of the reduced sample to FeO.

The reactivity rate (rinst) was defined as

where Δm is the weight loss during time step Δt and mtis the nonreacted sample mass at that instant of time.

The solid conversion during reduction(Xred)in the TGA was calculated as:

where m0is the initial sample mass.mtis the non-reacted sample mass at that instant of time.

2.3.Batch fluidized bed reactor

A batch fluidized bed reactor was used to evaluate the cyclic stability of the carriers,as well as the chemical looping performance with biomass.The CLG process,i.e.the cycling between a fuel reactor and an air reactor was simulated by periodically feeding solid fuel and oxidizing gas.Between each reduction and oxidation period the reactor was purged with N2for 5 min to avoid gas mixing.Fig.1 shows the schematic diagram of the reactor setup.It consisted of a gas supply system,a quartz fluidized-bed reactor,a solid recovery filter and a gas analysis system.The fluidized bedreactor,with a length of 800 mm and an inner diameter of 32 mm,was heated by an electrical furnace and the temperature was measured by a K-type thermocouple.A porous quartz plate located at 350 mm from the bottom of the reactor was used as sample holder and gas distributor.The gas supply system was consisted of different mass flow controllers and steam generator.A constant flowtype pump (TBP-50A) was used to feed an accurate volume of water to a steam generator,where the water was vaporized.Furthermore,the inlet flow was kept at 3 L?min-1during all of the batch experiments.

Fig.1.Schematic diagram of the fluidized-bed reactor.

In the cycle stability test,approximately 30 g of carrier particles were performed at 900 °C for 50 redox cycles.The reduction step was carried out with 15% CO in N2for 10 min,while 5% O2in N2was used as oxidizing agent for 10 min.Between the reducing and oxidizing periods,the reactor was purged with N2for 5 min.The elutriated solids with a size lower than 45 μm were considered as generated by attrition[3].In addition,the samples after the 50th redox cycling test were characterized.

The CO conversion,XCO,is calculated as:

Here,yi(i=CO,CO2) corresponds to the volume fraction of the component i in the flue gas.

The attrition rate (vattrition,%?h-1) is calculated as:where melutis the mass of the elutriated particles with a size lower than 45 μm.The particles of this size will have a short residence time in a commercial unit and thus be of little use in the process.mOCis the total mass of the oxygen carrier used.t is the operation time (h).

The average lifetime (h) is calculated as:

In the CLG of biomass test,oxygen carrier particles were placed on the porous plate and heated to the desired temperature in 5%O2.The mass of carrier particles depended on the supply oxygen coefficient (O/B),which defined as the molar oxygen supplied by oxygen carrier from Fe2O3to FeO to the molar oxygen needed by the biomass complete combustion.The O/B ratio selected was 0,0.5,1 and 1.5.After stabilization,the gas was switched to 30%steam in N2.During the reducing period,2 g of biomass was fed into the reactor.The reduction duration was set as 20 min.After reduction,5%O2in N2was introduced for 20 min to guarantee carrier complete oxidation.The outlet gas passed through a filter and condenser in sequence to remove fine ash and water.Then,the concentration of the outlet gas was analyzed by a NGA2000 type gas analyzer (EMERSON Company,USA) measuring 0.00–20.00%(volume) CO,0.00–100.00% (volume) CO2,0.00–10.00% (volume)CH4,0.00–25.00% (volume) O2and 0.0–50.0% (volume) H2.The concentrations of C2–C4hydrocarbons are analyzed by a gas chromatograph (GC).

Total volume flow(Fout)is calculated by the mass balance of N2flow introduced:

where yiis the volume fraction of the component i(CO,CO2,H2,CH4or C2Hm) in the outlet gas flow on a dry basis.

Carbon conversion is calculated by the mole of carbon in the flue gas to the mole of carbon in the fuel:

where NC,fuelis the mole of carbon fed to the reactor.

H2yield (YH2),CO yield (YCO) and syngas yield (Ysyngas) per unit of solid fuel fed can be calculated by:

3.Results and Discussion

3.1.Structural characterization of the carriers

Table 3 shows the main crystalline phases of the tested materials detected by X-ray diffraction (XRD) analysis.The XRD results revealed that all carriers were composed of Fe2O3,which was the main active phase.The reduction products considered in this paper were Fe3O4and FeO.Fe3O4was an intermediate state between Fe2O3and FeO [36].The continuous process of “Fe2O3→Fe3O4→FeO” could be decoupled into two reactions “Fe2O3→Fe3O4” and“Fe3O4→FeO”.The reaction could be simplistically described as the phase boundary reaction mechanism [36–41].For the steelmaking wastes,calcium oxide,calcium ferrite,magnesium oxide and magnesium ferrite were also detected.The interactions between Ca/Mg oxides and Fe2O3resulted in the formation of CaFe2O4and MgFe2O4,respectively.In the reduction period,MgFe2O4(MgO?Fe2O3) and CaFe2O4(CaO?Fe2O3) could be reduced to MgO?Fe3O4/MgO?FeO and CaO?Fe3O4/CaO?FeO,respectively[42,43].Besides,Ca2SiO4and CaAl2Si2O8were detected in LD slag and BF dust.

Table 3Phase compositions in the four oxygen carriers identified by XRD

Table 4Attrition rate and lifetime of the four oxygen carriers

The SEM images of the fresh carriers were presented in Fig.2.As can be seen,all the carriers exhibited porous structure.In Fig.2a and e (hematite),the fresh hematite particle was composed of irregular grains.The particle porosity was relatively high,which was beneficial to gas diffusion.In Fig.2b and f (BF slag),the initial sample displayed a rather loose and porous morphology.Dendrites-type grains randomly arranged on the surface.The fresh LD slag sample(Fig.2c and g)possessed a porous surface.The fresh BF dust sample(Fig.2d and h)possessed a spherical structure with dendrites-type grains arranged on the surface.

3.2.Reduction reactivity of steelmaking waste

The effect of reducing agent on the reduction reactivity of the four oxygen carriers was investigated using TGA (see Figs.3,4 and 5).For all oxygen carriers tested,the reducing agent in order of reducing reactivity was H2>CO>CH4,in spite of the concentration of H2was 3 times lower than CO or CH4.It is also observed that the reactivity rate initially increased significantly and then rapidly decreased with reaction time.This was because the decrease of available lattice oxygen for reduction with reaction time.Additionally,the maximum reactivity rate was obtained at the first minute for all tested carriers.

With regard to hematite,the oxygen transfer capacity was 8.32%,which was notably higher than that of steelmaking waste because of the higher Fe2O3content (75.41%,mass).Even so,the solid conversion of hematite with H2,CO or CH4was lower than that found for steelmaking wastes.Considering BF slag,the oxygen transfer capacity was 7.47%.This material showed a solid conversion of 74.9% with H2,while with CO and CH4was 53.0% and 13.6%,respectively.It was obvious that LD slag presented a lower weight loss with CO,H2or CH4when compared to the othermaterials.This can be attributed to the low Fe2O3content(19.27%,mass).The oxygen transport capacity of LD slag was only 4.77%.In spite of its lower oxygen transport capacity,LD slag presented a relative better reactivity rate,reaching a solid conversion of 76.64%,53.41% and 21.54% with H2,CO and CH4,respectively.For BF dust,the oxygen transfer capacity was 7.36%.This carrier seemed to be the best performing material,presenting the highest solid conversion with H2or CO.When reducing with H2,the weight loss was close to 7.15%,which was comparable to the theoretical weight loss of 7.36% corresponding to the reduction of Fe2O3to FeO.Additionally,the solid conversion was close to 97.7%.This observation also confirmed that the active content (Fe2O3) in BF dust was almost completely reduced to FeO.When CO was the reducing agent,the solid conversion could reach 82.5%.This value was around 2.12 times higher compared with hematite.In addition,the solid conversion for BF dust with CH4was 17.2%,which was slightly higher than hematite.

In order to further reveal the behavior of the four carriers,the reduction reactivity was investigated over a wide range of temperature using 15% CO in N2as reducing agent.Figs.6 and 7 present the reactivity rate and the solid conversion for various carriers at different temperatures,respectively.Unsurprisingly,the reaction temperature had a positive effect on the reaction rate:as the reaction temperature increased,the reaction rate also increased.As a result,the solid conversion for all carriers tended to increase with reaction temperature.As for hematite,solid conversion increased from to 33.3% to 44.9% over the temperature range 850–950 °C.For BF slag,the rise of temperature from 850 to 950 °C led to a slight increase of the solid conversion from 39.2%to 50.0%.Considering LD slag,solid conversion increased from 41.2% at 850 °C to 67.3% at 950 °C,which was 1.24–1.50 times higher than that for hematite.Notably,BF dust achieved the highest solid conversion among these materials at 850–950°C.This value in case of BF dust increased from 76.57% at 850 °C to 90.00% at 950 °C,which was 2.00–2.30 times in comparison to that of hematite.

Results from the TGA experiments indicate that steelmaking wastes,especially LD slag and BF dust,are the promising oxygen carriers in terms of reduction reactivity.Compared to hematite,steelmaking waste had a higher Ca-Mg content.The addition of Ca species into Fe-based carrier could increase the reduction rate of Fe2O3[22].As a consequence,the solid conversion was higher for the carrier with a higher Ca content.

3.3.Cyclic stability of steelmaking waste

The experiments of consecutive 50 redox cycles over hematite,BF slag,LD slag and BF dust were conducted using CO as reducing gas at 900°C in the fluidized bed reactor to evaluate the redox performance.Fig.8 shows the CO conversion with reaction time for the four carriers in the 1st and 50th reduction process at 900 °C.Both hematite and BF slag possessed excellent cyclic stability and slight decrease in CO conversion could be observed over 50 redox cycles.As for BF dust,however,the 50th cycled BF dust presented a lower reduction reactivity compared to the fresh one,revealing an obvious deactivation of BF dust over the 50 redox cycles.By contrast,LD slag showed a slightly higher CO conversion after 50 cycles,meaning a slight activation of LD slag.Moreover,LD slag also had the highest CO conversion at the initial period and then rapidly decreased with reaction time,which could be attributed to the low Fe2O3mass content (19.27%) in LD slag.

Fig.2.SEM images of fresh hematite (a and e),BF slag (b and f),LD slag (c and g),BF dust (d and h).

Fig.3.Effect of reducing agent on the weight loss of the four oxygen carriers at 900 °C.

Fig.4.Effect of reducing agent on the reactivity rate of the four oxygen carriers at 900 °C.

Fig.5.Effect of reducing agent on solid conversion of the four oxygen carriers at 900 °C.

Fig.6.Effect of reaction temperature on reactivity rate of the four oxygen carriers with CO.

Fig.7.Effect of reaction temperature on solid conversion of the four oxygen carriers with CO.

Fig.8.CO conversion with reaction time for the four oxygen carriers over the 50 redox cycles at 900 °C.

Table 4 shows the attrition rate and the corresponding lifetime of all the investigated carriers during 50 redox cycles in the fluidized bed reactor.Hematite appeared to have a higher lifetime(around 1270 h) compared to steelmaking wastes.The material BF slag showed a superior attrition index (0.094%.h-1),which was equivalent to lifetime of 1070 h.LD slag presented the lowest attrition rate of 0.081%.h-1among the three steelmaking wastes,which reflected a superior lifetime (1240 h).BF dust showed the highest attrition index(0.122%.h-1)among the tested carriers,corresponding to a lifetime of 820 h.Generally,particle attrition was caused by particle surface abrasion [44].This proved that BF dust showed the poor mechanical stability during the 50 redox cycles.

In order to analyze the microstructure evolution of the carriers,BET and SEM were used to characterize the samples after 50 redox cycles.Table 5 shows the BET surface area of the fresh and the carriers after 50 redox cycles.The BET surface area of hematite decreased from 0.872 m2.g-1to 0.671 m2.g-1after 50 redox cycles.Remarkably,the BET surface area values of the three steelmaking wastes were notably lower than hematite.The BET surface area of BF slag and LD slag decreased by 25.6% and 27.4% (from 0.893 to 0.664 m2.g-1and 0.858 to 0.645 m2.g-1),respectively;whilefor BF dust,its BET surface area decreased dramatically (from 0.736 to 0.217 m2.g-1,a 70.5% decrease) between the first and 50th cycle.

Table 5BET surface area of the four oxygen carriers (m2.g-1)

Fig.9 illustrates the SEM images of the oxygen carriers after 50 redox cycles.By comparing the morphology of the used carriers with the fresh one,there were drastic changes over the 50 redox cycles.Compared with the SEM image of fresh hematite,there were fine sintered particles deposited on the surface of the cycled sample.The surface of the 50th cycled BF slag was similar to that in the initial state.This corroborated the retention of multi-porous structure shown by this material during CLC with CO.Compared with the fresh LD slag,the surface of the used one tended to be denser and some fragments stuck on the surface.This was probably due to the attrition and slight agglomeration.On the contrary,the surface of the cycled BF dust appeared to be much coarser compared with the fresh one.The feature size of the grains on the surface became larger.The formation of larger grains could be attributed to grains coalescing and sintering.As for BF dust,the BET surface area significantly decreased with the number of redox cycles from 0.736 m2.g-1to 0.217 m2.g-1after the fiftieth redox cycle.The remarkable decrease of BET surface area for BF dust was in agreement with the change of SEM images during redox cycles.These results revealed that the change of particle morphology was responsible for the low reactivity of the 50th cycled BF dust samples.As a consequence,BF dust was not suitable to be used as oxygen carrier in chemical looping process.Thus,hematite,BF slag and LD slag were chosen for further investigation.

Fig.9.SEM images of hematite (a′ and e′),BF slag (b′ and f′),LD slag (c′ and g′),BF dust (d′ and h′) after 50 redox cycles.

3.4.Reactivity of steelmaking waste in CLG

The reactivity of steelmaking waste was evaluated in the fluidized bed reactor for CLG of biomass.The main reactions involved in CLG of biomass were summarized in the following reactions.When biomass was added into the reactor,biomass pyrolysis(R1)occurred simultaneously with tar cracking(R2)and char gasification (R3)–(R4).Additionally,hydrocarbon reforming (R5)–(R6)could occur.Subsequently,Fe2O3in oxygen carrier reacted with gaseous products (H2,CO or CH4) following reactions (R7)–(R9).

The effects of the O/B ratio on gas concentrations,syngas yield and carbon conversion for hematite,BF slag and LD slag were investigated at 900 °C,and the results were shown in Figs.10 and 11.From Fig.10,it can be seen that the main components from CLG of biomass were CO,CO2and H2,along with small amounts of CH4and C2Hm.For all carriers,the concentration of CO2tended to increase with the O/B ratio,while both CO and H2displayed the opposite trends.Besides,the concentrations of both CH4and C2Hmslightly decreased with the O/B ratio.Unsurprisingly,higher O/B led to more amount of active lattice oxygen available for combustion per unit of biomass,resulting in higher consumption of CO and H2.In addition,more oxygen carrier inventory could enhance tar cracking reaction to generate more gaseous products.

Fig.11 shows the variation of H2yield,CO yield and carbon conversion as a function of O/B ratio for the four oxygen carriers at 900°C.With the increase of O/B ratio from 0 to 1.5,the carbon conversion significantly increased from 68.45% to 81.68%,68.45% to 81.77% and 68.45% to 83.45% for hematite,BF slag and LD slag,respectively.In addition,LD slag showed the highest carbon conversion,followed by BF slag.More importantly,H2yield was higher than CO yield for all oxygen carriers.H2/CO ratio decreased with the increase of the O/B ratio.This can be explained as follows:(1) The active lattice oxygen available for conversion per unit of biomass increased with the O/B ratio;(2) The reduction rate of Fe2O3with H2was dramatically higher than that with CO.Furthermore,both of the H2and CO yields initially performed with a gradually increasing tendency(0

From the results above,it can be concluded that the presence of oxygen carrier was beneficial for increasing carbon conversion as well as syngas yield.On the one hand,the presence of oxygen carrier could promote biomass pyrolysis (R1),tar cracking (R2) and char gasification (R3)–(R4) and hydrocarbon reforming (R5)–(R6),thereby increasing carbon conversion.On the other hand,H2 and CO could also react with oxygen carrier (R7)–(R9),subsequently decreasing syngas yield.Consequently,it was crucial to reasonably control the O/B ratio.After comprehensive consideration of CLG performance and oxygen carrier stability,an O/B molar ratio of 1 was chosen for further exploration.

Fig.10.Effect of O/B ratio on the proportion of the main gaseous products for hematite,BF slag and LD slag at 900 °C.

Fig.11.Effect of O/B ratio on gas yield and carbon conversion for hematite,BF slag and LD slag at 900 °C.

Temperature was a critical parameter for chemical looping gasification.The effects of reaction temperature (800–950 °C) on H2yield,CO yield and carbon conversion at the O/B ratio of 1 were displayed in Fig.12.It was observed that temperature had a significant effect on carbon conversion.As the temperature increased from 800 to 950 °C,the carbon conversion dramatically increased from 68.13% to 80.71%,69.32% to 80.78% and 70.40% to 82.68%for hematite,BF slag and LD slag,respectively.Increasing reaction temperature could mainly promote the biomass pyrolysis(R1),tar cracking (R2) and char gasification (R3)–(R4) and hydrocarbonreforming (R5)–(R6).Besides,the reduction reactivity of oxygen carrier could increase.

Fig.12.Effect of temperature on gas yield and carbon conversion for hematite,BF slag and LD slag at the O/B ratio of 1.

As regards gas yield,the yields of both H2and CO had a slightly increasing tendency with the temperature from 800 to 850°C,and subsequently marginally changed when temperature reached 900 °C.Further increasing temperature from 900 to 950 °C could significantly reduce H2and CO yields.It can be concluded that the reaction temperature between 850 °C and 900 °C were an acceptable range.To ensure a higher carbon conversion,the reaction temperature of 900 °C was chosen as the typical operating condition.

Furthermore,LD slag showed the highest syngas yield and carbon conversion compared to the other carriers (hematite and BF slag).In the typical condition(O/B=1,T=900°C),the syngas yield was 0.527 m3?kg-1,0.537 m3?kg-1and 0.548 m3?kg-1for hematite,BF slag and LD slag,respectively.Generally,there could be two mechanisms for the higher gas yield of LD slag.The first one was the catalytic effects of alkaline metals (Ca and Mg) in steelmaking waste on biomass gasification.According to the results in Table 1,the amount of CaO and MgO in steelmaking waste was significantly higher than that in hematite.This value for LD slag was 17.98% and 35.1%,respectively.The catalytic effect of Ca and Mg in terms of volatiles reforming and cracking of tar into light gases was significantly greater Fe [19,29].The other was the higher reduction reactivity of steelmaking waste compared to hematite.As previously mentioned,the presence of Ca and Mg in steelmaking wastes could improve the reduction reactivity rate of Fe2O3.Thus,it can be anticipated that LD slag will be a promising oxygen carrier for chemical looping application.

4.Conclusions

In this study,Ca-and Mg-rich wastes,such as blast furnace slag(BF slag),blast furnace dust(BF dust)and Linz-Donawitz converter slag (LD slag),from steelmaking industry were demonstrated as oxygen carriers for chemical looping process.These materials contained sufficient active iron oxides and proper contents of Ca or Mg.The application of these materials as oxygen carriers in chemical looping process was experimentally evaluated.

The reduction experiments verified that Ca-and Mg-rich steelmaking waste,especially LD slag and BF dust,showed a higher reducibility compared to hematite.This indicated that steelmaking waste was a promising candidate as an oxygen carrier for chemical looping process.In addition,all carriers expect BF dust showed a well-preserved stability during long-term operation.

As regards the CLG behavior of Ca-and Mg-rich wastes,the effects of O/B and temperature on syngas yields were evaluated.The results indicated that a higher syngas production was achieved Ca-and Mg-rich waste compared to hematite.The presence of Ca or Mg in steelmaking waste could improve both reduction rate of Fe2O3and gasification rate of biomass.The syngas yields initially performed with a gradually increasing tendency with the increase of O/B ratio or reaction temperature and then subsequently decreased.Under the optimum condition (O/B=1,temperature=900 °C),the syngas yield achieved was 0.548 m3?kg-1using LD slag as oxygen carrier.These findings indicated that LD slag was a promising oxygen carrier for chemical looping process in terms of perfect reducibility and superior CLG behavior.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

This research was supported by the National Natural Science Foundation of China (Grant No.52006104),and the Fundamental Research Funds for the Central Universities (No.30919011237).